Bravo
If ARM was an arm, BRN would be its biceps💪!
Hi Bravo,
As always, I was ready to apply the wet blanket, and I would ov too, until Sean disclosed our algorithm product line:
US2022210586A1 AUTOMATIC TINNITUS MASKER FOR AN EAR-WEARABLE ELECTRONIC DEVICE 20220630 Starkey
View attachment 70667
[0075] … the controller comprises, or is operatively coupled to, a processor configured with instructions to classify, via a first neural network, the acoustic environment of the wearer as a specified one of a plurality of disparate acoustic environments, and process one or more of the physiologic sensor signals, the non-physiologic sensor signals, and the contextual factor data, via a second neural network, to adjust the tinnitus masking sound produced by the sound generator using the one or more of the physiologic sensor signals, the non-physiologic sensor signals, and the contextual factor data, and parameter values associated with the specified acoustic environment.
[0080] Example Ex44. The device according to Ex43, wherein the neural network comprises one or more of a deep neural network (DNN), a feedforward neural network (FNN), a recurrent neural network (RNN), a long short-term memory (LSTM), gated recurrent units (GRU), light gated recurrent units (LiGRU), a convolutional neural network (CNN), and a spiking neural network.
[0095] In accordance with any of the embodiments disclosed herein, the controller 120 can include, or be coupled to, a machine learning processor 124 configured to execute computer code or instructions (e.g., firmware, software) including one or more machine learning algorithms 126 . The machine learning processor 124 is configured to process one or more of the physiologic sensor signals, non-physiologic sensor signals, microphone signals, and contextual factor data via one or more machine learning algorithms 126 to detect one or more of presence, absence, and severity of tinnitus of the wearer of the hearing device 100 . Sensor, contextual factor data, and/or wearer input (e.g., manual overrides) received by the machine learning processor 124 are used to inform and refine one or more machine learning algorithms 126 executable by the machine learning processor 124 to automatically enhance and customize tinnitus detection and mitigation implemented by the hearing device 100 for a particular hearing device wearer.
I don't know if I'm more shocked at the promising content of this patent or that I avoided getting another dose of that wet blanket of yours.